Matches in SemOpenAlex for { <https://semopenalex.org/work/W2777654136> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W2777654136 abstract "We introduce scGAN, a novel extension of conditional Generative Adversarial Networks (GAN) tailored for the challenging problem of shadow detection in images. Previous methods for shadow detection focus on learning the local appearance of shadow regions, while using limited local context reasoning in the form of pairwise potentials in a Conditional Random Field. In contrast, the proposed adversarial approach is able to model higher level relationships and global scene characteristics. We train a shadow detector that corresponds to the generator of a conditional GAN, and augment its shadow accuracy by combining the typical GAN loss with a data loss term. Due to the unbalanced distribution of the shadow labels, we use weighted cross entropy. With the standard GAN architecture, properly setting the weight for the cross entropy would require training multiple GANs, a computationally expensive grid procedure. In scGAN, we introduce an additional sensitivity parameter w to the generator. The proposed approach effectively parameterizes the loss of the trained detector. The resulting shadow detector is a single network that can generate shadow maps corresponding to different sensitivity levels, obviating the need for multiple models and a costly training procedure. We evaluate our method on the large-scale SBU and UCF shadow datasets, and observe up to 17% error reduction with respect to the previous state-of-the-art method." @default.
- W2777654136 created "2018-01-05" @default.
- W2777654136 creator A5007816191 @default.
- W2777654136 creator A5016908947 @default.
- W2777654136 creator A5038957718 @default.
- W2777654136 creator A5047566533 @default.
- W2777654136 creator A5076307452 @default.
- W2777654136 date "2017-10-01" @default.
- W2777654136 modified "2023-10-15" @default.
- W2777654136 title "Shadow Detection with Conditional Generative Adversarial Networks" @default.
- W2777654136 cites W1903029394 @default.
- W2777654136 cites W1998270967 @default.
- W2777654136 cites W2027850463 @default.
- W2777654136 cites W2073839959 @default.
- W2777654136 cites W2095761827 @default.
- W2777654136 cites W2113449189 @default.
- W2777654136 cites W2122514671 @default.
- W2777654136 cites W2123302273 @default.
- W2777654136 cites W2141562793 @default.
- W2777654136 cites W2147558395 @default.
- W2777654136 cites W2147800946 @default.
- W2777654136 cites W2166502676 @default.
- W2777654136 cites W2206865673 @default.
- W2777654136 cites W2470948946 @default.
- W2777654136 cites W2544806203 @default.
- W2777654136 cites W2963420272 @default.
- W2777654136 cites W3004335748 @default.
- W2777654136 doi "https://doi.org/10.1109/iccv.2017.483" @default.
- W2777654136 hasPublicationYear "2017" @default.
- W2777654136 type Work @default.
- W2777654136 sameAs 2777654136 @default.
- W2777654136 citedByCount "133" @default.
- W2777654136 countsByYear W27776541362017 @default.
- W2777654136 countsByYear W27776541362018 @default.
- W2777654136 countsByYear W27776541362019 @default.
- W2777654136 countsByYear W27776541362020 @default.
- W2777654136 countsByYear W27776541362021 @default.
- W2777654136 countsByYear W27776541362022 @default.
- W2777654136 countsByYear W27776541362023 @default.
- W2777654136 crossrefType "proceedings-article" @default.
- W2777654136 hasAuthorship W2777654136A5007816191 @default.
- W2777654136 hasAuthorship W2777654136A5016908947 @default.
- W2777654136 hasAuthorship W2777654136A5038957718 @default.
- W2777654136 hasAuthorship W2777654136A5047566533 @default.
- W2777654136 hasAuthorship W2777654136A5076307452 @default.
- W2777654136 hasConcept C106301342 @default.
- W2777654136 hasConcept C11413529 @default.
- W2777654136 hasConcept C117797892 @default.
- W2777654136 hasConcept C120665830 @default.
- W2777654136 hasConcept C121332964 @default.
- W2777654136 hasConcept C152565575 @default.
- W2777654136 hasConcept C153180895 @default.
- W2777654136 hasConcept C154945302 @default.
- W2777654136 hasConcept C15744967 @default.
- W2777654136 hasConcept C184898388 @default.
- W2777654136 hasConcept C192209626 @default.
- W2777654136 hasConcept C31972630 @default.
- W2777654136 hasConcept C41008148 @default.
- W2777654136 hasConcept C542102704 @default.
- W2777654136 hasConcept C62520636 @default.
- W2777654136 hasConcept C76155785 @default.
- W2777654136 hasConcept C94915269 @default.
- W2777654136 hasConceptScore W2777654136C106301342 @default.
- W2777654136 hasConceptScore W2777654136C11413529 @default.
- W2777654136 hasConceptScore W2777654136C117797892 @default.
- W2777654136 hasConceptScore W2777654136C120665830 @default.
- W2777654136 hasConceptScore W2777654136C121332964 @default.
- W2777654136 hasConceptScore W2777654136C152565575 @default.
- W2777654136 hasConceptScore W2777654136C153180895 @default.
- W2777654136 hasConceptScore W2777654136C154945302 @default.
- W2777654136 hasConceptScore W2777654136C15744967 @default.
- W2777654136 hasConceptScore W2777654136C184898388 @default.
- W2777654136 hasConceptScore W2777654136C192209626 @default.
- W2777654136 hasConceptScore W2777654136C31972630 @default.
- W2777654136 hasConceptScore W2777654136C41008148 @default.
- W2777654136 hasConceptScore W2777654136C542102704 @default.
- W2777654136 hasConceptScore W2777654136C62520636 @default.
- W2777654136 hasConceptScore W2777654136C76155785 @default.
- W2777654136 hasConceptScore W2777654136C94915269 @default.
- W2777654136 hasLocation W27776541361 @default.
- W2777654136 hasOpenAccess W2777654136 @default.
- W2777654136 hasPrimaryLocation W27776541361 @default.
- W2777654136 hasRelatedWork W1497780831 @default.
- W2777654136 hasRelatedWork W1999949837 @default.
- W2777654136 hasRelatedWork W2034462085 @default.
- W2777654136 hasRelatedWork W2041408572 @default.
- W2777654136 hasRelatedWork W2096492340 @default.
- W2777654136 hasRelatedWork W2317337758 @default.
- W2777654136 hasRelatedWork W2357262985 @default.
- W2777654136 hasRelatedWork W3006419044 @default.
- W2777654136 hasRelatedWork W3014708372 @default.
- W2777654136 hasRelatedWork W3177406559 @default.
- W2777654136 isParatext "false" @default.
- W2777654136 isRetracted "false" @default.
- W2777654136 magId "2777654136" @default.
- W2777654136 workType "article" @default.